Psychiatry & Psychotherapy Podcast - IS ELON MUSK'S NEURALINK A SCIENCE FICTION HORROR SHOW OR THE SALVATION OF HUMANITY?

Episode Date: August 21, 2019

One of the more controversial components of the Neuralink presentation was Musk's inclusion of his beliefs about the future of humanity and artificial intelligence. During the press release he stated one of his goals was to create the ability to achieve a "full symbiosis with artificial intelligence," essentially removing the "existential threat of AI" which he believes will one day "leave us behind" (Neuralink, 2019). This goal has been met with a bit more skepticism, especially by the national media, than the medical applications of Neuralink's BMI. Forbes describes it as "a bit more fantastical" than the company's primary goal of treating brain disorders (Knapp, 2019). Other publications have been far less kind, such as The Atlantic, which published its coverage of the Neuralink press release with the title: "Elon Musk's Next Wild Promise: If someone is going to revolutionize what it means to be human, do we want it to be a tech titan?" (Mull, 2019). Although the New York Times surmised that "one of the biggest challenges may be for his scientists to match his grand vision," (Markoff, 2019), it serves as a good example of what most major media outlets have chosen to do: stick to the facts.   By listening to this episode, you can earn 1 Psychiatry CME Credits. Link to blog. Link to YouTube video. Engage in the dialogue on Dr. Puder's Instagram, Facebook, Twitter  

Hello and welcome to the psychiatry and psychotherapy podcast, with over 32,000 mental health professionals listening every episode. Why? Because we need to stick together to survive the mental health field. I'm here to talk about getting rid of burnout, increasing job satisfaction, and feeling like an expert in what you do. On today's episode, Dr. Cummings and I will be talking about brain machine interfaces, also called brain computer interfaces and mind machine interfaces. recently Elon Musk talked about this in a neuralink announcement in which he talked about this new device that they're going to put into the brain with a thousand connections that can go both in and out of the brain, meaning that it can communicate with your neurons and your neurons can communicate with potentially a machine, a computer.

And in this podcast episode, we're going to dive into some of the history of things that have already been put in the brain, like deep brain stimulators, we're going to talk about trans magnetic cranial stimulation, which gives us knowledge about how, you know, targeting different sites of the brain are going to help in things like depression and OCD. We're also going to be talking about specific pathways, which in the future we may target with machines like this,

OCD, depression, and schizophrenia, so we'll go through specific pathways and how we can target those in different diseases. Dr. Cummings and I have no conflicts of interest, either with psychopharm companies or with, you know, any of these companies like Neurrelink. And we'll also talk a little bit about other companies that are competing for this spot, so it's not just about one company. I really hope you find this episode interesting. and it was it was truly an amazing conversation to have with dr cummings it really did take me to the to kind of like the extent of my intellect it felt like so i hope you find it engaging before we start i want to make an announcement and the announcement is in part why i do this podcast when i wrote down my purpose when i

first started this podcast, I wrote down that I wanted to impact future providers of mental health. I wanted to give them a glimpse into the field, inspire the next generation of mental health professionals. I wanted to help them improve their empathy, their psychological safety, their ability to build a therapeutic alliance, their search for honest truth, their desire to read and to learn and to grow. and the announcement today is that I've started a Patreon account, which is a way to support this podcast financially through a monthly contribution. It can be small as $1 a month to around $50 a month, somewhere around there. And I'm looking for people who are seeing that this has a potential to have an high impact factor on the future

of mental health and those who want to support it. And, you know, I get these messages for medical students and residents that, you know, they've binged on all the episodes once or twice, and they feel like they've grown in their empathy, they've grown in their wisdom, they've grown in their knowledge, and a lot of them just say, I've grown in my excitement for the field. And so if you would like to support that, if that would be something that would be important for you, that would give you a sense of meaning and purpose, then I would really appreciate your support.

And up to now, I have not been receiving any financial support from any drug companies or any other sort of outside entities. And I've been self-funding, and I'm looking to basically find some partners and at different levels of support through Patreon, you can get different things. For example, at $10, you can get a shirt. At $25 a month, you can get access to a closed Facebook group

to dialogue and to ask questions and to meet other people who are into this thing. And at $50 a month, you can get access to the continued medical education that can go with this podcast. So at the end of the day, I'm hoping to impact the first, field of mental health as a whole and to make people's lives not just not depressed or not just

you know not having negative symptoms but really thriving and I hope that I can continue to do this for years and I think partnering with people will make that happen so please consider that and I will have a link to the Patreon and the show notes and on my website as well. Okay welcome back to the I am here once again with Dr. Cummings, an amazing psychiatrist who is currently working actually on a book on severe mental illness. So he's like kind of at the forefront of psychopharmacology and brain science and how this all interplays. But today what we are going to talk about is a recent press release by Neurrelink on brain machine interfaces. And the brain has 86 billion neurons, some with 10,000 connections. So there's 1,000 trillion

synaptic connections. And it's about 1 trillion bits per second, you know, processor that we have in our brain. The brain has about 1 to 1,000 terabytes of memory. You know, and considering that the U.S. Library of Congress only has 19 million volumes, which is about 10 terabytes of data. So our brain has one to one thousand terabytes of memory in different in different analysis they figure that out so today what we're going to talk about is some of the history of what has already been put in the brain so we'll talk about deep brain stimulators and for depression and for parkinson's and then we'll go on to the the the neuralink press release Elon musk and some of the stuff that they talked about in there that's new and also what's been done previously in the

research to date with like high density EEGs where they're trying to decode verbal things. And what we will then go into is we will talk about how this potentially could help people with mental health issues. What areas of the brain might need to be targeted for mental health and we'll go through different diseases. So Dr. Cummings, welcome. Thank you. I am pleased to be back.

Indeed, this touches on both my current life as a psychiatrist and pharmacologist as well as my previous existence as a physicist, because I'm always been interested in computers and information processing. When we start with the biology, biological systems have been processing information for a very long time, starting with very primitive systems, some of which we. still don't understand. For example, at Cornell University, they repeated the famous and classic Pavlovian experiment, except instead of dogs, which are very similar to us in terms of central nervous system, they did so with pea plants, which, as far as we know, don't actually have a nervous system. They paired a light and a fan. The fan by itself produced no response from the plants.

pairing a blue-green light with the fan, however, produced conditioning in which eventually the pea-plants would grow toward the fan without the light present, suggesting that in some way, even at a plant level, the plant was able to store memory that moving toward a particular stimulus which had been neutral, had potential positive benefits. that speaks to information and information processing being very old in biology. Biological systems like us are electrochemical in nature. We are different from computers in a very significant way in that computers process information so far in a very linear manner. There's an input, there's a CPU, there's an output. The human brain and the 86 billion neurons that were described. behave much more like an ant colony or a B colony

in which there is no a priori linear leader, if you will. Instead, different assemblies of neurons do different things. For example, the fusiform gyrus recognizes faces, the hippocampus reconstructs memory and lays down memory. The ventrometrial prefrontal cortex is involved in judgment and making estimates of how dangerous or not dangerous something is, that has been difficult to interface with computer systems because of the very different structure of information processing.

People have been making progress, though, and indeed with the advent of neuroimaging and functional MRI and PET scanning, we've now begun to identify specific circuits that may be abnormal, in some conditions like Parkinson's disease or in major depression or in some cases refractory hallucination. And indeed, one of the earliest interventions has been implantation of electrodes to either stimulate or inhibit a particular neural circuit. And that has produced benefits across a wide range of diseases involving particularly motor systems, because they tend to be more inherently linear in structure. Yeah, so how many electrodes are actually going into the brain

and some of these deep brain stimulators that are already there? The number ranges from about 5 to 20, and as in most such neurosurgeries, the patient is awake during the placement, and when they essentially embed an array of electrical, They can then activate each microelectrodes. These are tiny, much thinner than a human hair. They can activate each one individually and have the patient give them feedback about what has changed or feedback about, you know, they can have them attempt to move and see if the movement is smoother, better, easier. And when they find the ones that appear to be most beneficial, they can then program the control.

chip to activate those and not the other electrodes. And after such surgery, typically anywhere from two to four electrodes will be in the active state. And of course, they can then externally change the programming of the chip if they need to and can alter the treatment response. So that's been our first venture into altering the way the brain works by interfacing essentially microchip technology with the biology of the human brain. And how successful has this been for depression, Parkinson's?

In many cases, it's been very dramatic because usually the cases have been selected based on failure to respond to older, more broad treatments. In other words, psychopharmacology, people who are refractory, to other treatments sometimes have wound up in these studies because it's a bit more of a direct approach to modifying the way the brain is functioning. One of the limitations we have with almost all of our pharmacological agents is we're attempting to influence the way these computing elements

called neurons work by perturbing them repeatedly kind of at the outside or at the cell membrane, and then hoping that kind of produces changes in second messengers and in DNA expression that will alter the way that neuron and its companions behave. It's kind of a very remote approach, and for some illnesses, an ineffective approach. So one of the promises held by perhaps more direct interfaces is that there will be an ability to more directly influence what various brain circuits do. And if it's a circuit that's misbehaving, either underactivity or overactivity, that can be changed by a microprocessor that senses whether the activity level is what's desired

or it's not enough or it's too much. So within the deep brain stimulation and depression, what part of the brain are they targeting Typically, they're targeting, as we've done with rapid transcranial magnetic stimulation, usually it is the anterior portion of the limbic circuit, the circuit of pepess in the dominant lobe, and indeed stimulation of that portion in most people produces an improvement in mood. One of the things that we figured out a number of years ago was that, of course, have two limbic systems, one in the dominant hemisphere, one in the non-dominant. Those normally function in fairly good balance with each other.

One of the things of many that may be wrong in refractory major depression is that those two limbic systems no longer communicate adequately, no longer function in unison. And indeed, if the dominant hemisphere limbicicic system, system is underactive, the person exhibits a lot of the signs and symptoms of major depression. Certainly a lot of that came out of stroke research where people actually had strokes involving that portion and then became depressed. So by dominant you mean? Oh, and most people, this would be the left hemisphere of the brain if you're right-handed.

The odds are 99% that your left hemisphere is dominant, meaning that's where your language centers are. if you are left-handed, it's a more mixed picture. You're still 60% likely to be left hemisphere dominant, but you may be right hemisphere dominant, or you may have a mixed dominance. So TMS, they're using magnets to stimulate that part of the brain. And with the deep brain stimulators,

they're also trying to stimulate that same part of the brain. Yeah. In essence, the end product is, the same, even though the approach is very different. Rapid transcranial magnetic stimulation depends on rapidly reversing an intense magnetic field which induces electrical currents in the brain tissue underlying the magnet, whereas, of course, the electrodes directly apply an electrical differential to brain tissue. One other question I had on deep brain stimulation is, what are the side effects?

Typically, it has been pretty well tolerated. Of course, this is a major neurosurgery, so there are all of the risks and hazards of having essentially a neurosurgery done. However, the overall safety record of deep brain stimulation has been quite good. They do have cases where they try, they put electrodes in place, and either they haven't been able to get the right placement or the pathology is more complicated. anatomically than they thought, and it's not effective. But the rate of bad outcome, meaning something catastrophic happens because of the implant is almost nil.

So, you know, are the current electrodes they're placing into people rigid? No, they are flexible. They are tiny. And do not involve a lot of destruction of neural tissue. The brain is unlike indeed computers. The brain is very soft. In fact, if your brain was not supported by cerebral spinal fluid and the pressure inside blood vessels,

and you were to take your brain out and put it on a desktop, it would basically just pancake. It would flatten out. It's about as resilient on its own as very thick shaving cream. So have there been like an immune response to these probes in the brain? Typically no. I mean, I wouldn't say that categorically because, of course, if you do something to enough people, you'll find somebody eventually who is reactive.

However, these probes have been made out of materials that are biologically compatible, that are not prone to cause an inflammatory response or reaction. So in Elon Musk's neuralink press release, he talks about, you know, creating basically a brain machine interface, which has not, you know, only like 10, but has potentially 10,000 electrodes. This will be inserted through a robotic machine very precisely to avoid putting them anywhere near blood vessels. and they have already put this in monkeys. It leaked out to the point that the monkey could control a keyboard. Yes. And when he talked about it in the beginning,

there's a couple things I noticed. One is he was sweating profusely, and he's very nervous. And I think that Elon Musk really thinks that AI, artificial intelligence, is going to take over the world. And there's, I think he really believes that the only way to combat AI is to have a symbiotic relationship with AI in the human brain? I'm not sure that AI is going to take over the world.

That is, I'm not sure that we're going to wake up tomorrow, and SkyNet and Terminator are going to be real things. It is true that we've been interacting with technology basically since we first made a tool. and certainly these days with artificial intelligence, we are shaping it via programming and via training, and AI is also shaping us. I was recently at a conference where one person looked at intelligence quotient scores over the last 20 years and found in a fairly large group about 5,000 people that visual intelligence, that is the ability to interpret symbol, graphs, et cetera, has actually increased on average by about four points or about a third of a standard deviation. In contrast, verbal intelligence has declined by about seven points or almost half a

standard deviation in the same group. And that may speak to the issue that right now our main interaction with artificial intelligence is either via voice or much more commonly via a screen as in smartphones, computer screens, tablets, etc. Yeah, I do find that reading and long-form reading is not as popular, especially with the youth, you know, YouTube, podcasts. I think that it's coming into this niche because it's easier to consume information through these methods. And I think something else, well, I'm speaking for myself here, those of us who are old,

remember the days before things talked back. These days, of course, young children growing up or growing up in a world in which objects and appliances and so forth are capable of responding to them verbally and will understand what they say verbally. I don't know if this story is apocryful or not, but one of these stories told by this lecture at this particular conference was that his three-year-old was complaining that the family

toaster was broken because when she spoke to it, it did not respond. Hmm. That's really something. So, you know, I think Elon Musk, in the psychology of Elon Musk, you know, he's fearing human apocalypse, one, by, you know, environmental means, which is why I think he's so passionate about electronic vehicles. and solar power. And then the second thing, I think, is he's thinking AI is potentially a huge danger. But I think for the topic of our discussion, well, go ahead. You were going to say something.

I was going to say that, you know, I think if AI were to become, well, it's like all technologies, technologies can, of course, be used for good or bad things. But that's largely a product of what we choose to do with. of them. Whether AI will itself eventually have an opinion about what it should or should not do, I think will depend on whether AI systems ever achieve sentience. It's not clear that they will. It's also not clear that they won't. There are proponents on both sides. Curswile, Dr. Curswile, who invented the Xerox machine, firmly believes that eventually computers will become sentient, while other people have grave doubts about them

becoming self-aware, conscious, that sort of thing. Yeah, I personally think that AI is already being used and probably will continue to be used for discrete tasks. Things like having your car drive you to work instead of you driving to work, I think that is like in the next couple years. trucks being you know self-driving from one destination to another

I think we're a couple years away from that yeah that was disclosure my wife drives a Tesla and indeed it on the freeway from on ramp to off ramp it will drive you from getting on the freeway to getting off the freeway and that will turn

as it exits the freeway it will turn control back over to the human wow and we have no financial interests. Neither of us with any pharmaceutical companies or any mind-brain devices. I'd like to put that out there as well. Now, commenting further on Tesla and also on Elon Musk, I mean, one of the interesting things about his Neurrelink company, this is another company outside of Tesla or SpaceX that he founded in 2016.

that is essentially dedicated to pursuing a more efficient brain computer interface. And certainly, I think as things progress, we probably will develop systems that will be able to better interpret input from the brain directly. And we'll be able to respond to things that the brain may be doing correctly or incorrectly in a much more elegant manner than our current systems do. Yeah, and I think it's like speed. So the speed at which we can take in data is limited by our vision and our hearing.

The speed that we can put out information is limited by our speech or by, you know, as fast as we can type. And I think that what he was hinting at and what he's hinted at in other talks is that what he wants to do is drastically increase the speed that both occur.

And he's hoping that the brain interface that he's developing is capable of that. Now, there are a lot of other companies that are racing to do this IBM, John Hopkins, MIT, Facebook. Facebook has a speech decoding device. There's some studies that they're teaming up with different universities on. and it's a non-invasive way. And then you have DARPA and BrainGade and Apple and Google. And there's another company, Kernel, and they're trying to create algorithms that can make the neurons exchange data

wirelessly through like a hat-light receiver. So there's all these companies right now racing to create these like, you know, mind-brain interfaces. Yeah. There are a variety of people pursuing a number of different people. pathways to see if we can actually reach a point where there can be actual exchange of discrete information between neural systems and computer chips because, of course, the advantage that a computer system has is speed. While we have advantages in terms of parallel processing of information, decentralized processing of information, computers can

process information much more rapidly than human beings can. And if eventually you had a human computer hybrid in terms of mental processes, you'd have a smarter, faster, human being. So coming back to your sort of takeaway from the presentation and the paper that Elon Musk put out, Do you have any other thoughts, any other sort of innovations that he spoke about that are worthy to discuss? I think probably the main area of interest in this is he's obviously dedicated to pursuing a very large investment in this area. And in that sense, the research and development, I think will likely reach the threshold where discrete information can be, processed, perhaps outside the brain and then re-entered into the brain so that there becomes a two-way communication path between various brain circuits and computer components.

That holds a lot of promise for the future in terms of treating both mental and neurologic diseases because at their heart, all neurologic and mental diseases involve malfunctioning. or misfunctioning or damage to discrete neural circuits or neural assemblies. And if we can figure out ways to correct those errors, we may be able to conquer some of the illnesses that so far have alluded us. Yeah, I think the thing that gets me excited is the potential to combat things like dementia, potentially combat things like, you know, severe, severe depression in a more exact way.

And frontal lobe damage, some of these areas in mental health where it's like we don't have a lot of options for these people. Yeah. Well, I think a lot of mental health will come from two separate directions in terms of help.

One is the technology because indeed for a lot of functions, once we have an interface that is truly functional, you can improve or tune up what is there. The other pathway I think that may help us in the long run is as we learn more and more about how to manipulate relevant growth factors in the brain, we may be able to actually repair some of the biological damage

that our illnesses involve. In fact, one of the reasons we do so badly with a number of illnesses is that often the pathology has progressed for years or decades before we get to the point of introducing treatment. And often that involves chronic inflammation, cell loss, loss of synaptic connections, and both the technology approach and the biological approach, particularly if used together may be able to help reverse some of those core pathologies. So one of my thoughts when I was thinking about this was that, you know,

have like 86 billion neurons, some of 10,000 connections. And, you know, even at the level of sophistication that they were looking at, they're talking about 10,000 pathways out of the brain or into the brain. Do you think that that's potentially enough? One of the critical elements that will be, no, I don't think that's enough. Numerically, of course, 10,000 compared to 86 billion is poultry. number but i suspect the number will increase as the technology improves um an area we haven't touched on um that may help this as well is there are also research lines going on in developing

micro robots that may be able actually to uh rather than doing perhaps a gross open head surgery you could introduce nanobots to build structures inside biological systems. And instead of 10,000 connections, you might wind up with millions of connections at critical points. I think the other thing that will help that, of course, is as we learn more and more about the micro anatomy and histology of brain systems, all information processing systems have critical points where information is either being sent out or coming in and it's modulation of things usually at those critical points where you can alter overall system functioning quite a bit.

I misspoke. It's not 10,000, but what he was talking about was that each chip is about 1,000. And so if you had 10 chips, that could be 10,000. So let me let me, um, Clarify that. Yeah. But I think that approach eventually will be supplanted by a more elegant perhaps nanorobot architecture. Because I mean you can program very microscopic robots basically to form chips in biological systems.

It's actually been done. And in fact, there are some nanorobots that are now have reached the point. where they can use ATP or GTP as a power source, which of course makes them compatible with biological systems in terms of you don't need a battery. The biological system will actually provide power for these nanorobots. And we're talking about things that are two, three microns in size

such that you could have literally billions of them infused, and they could be pre-programmed. to accomplish specific tasks. You said that this has been done? It's been done in a very primitive form. For example, there are nanorobots that have been developed to help identify specific collections of cholesterol and other molecules in blood vessels.

And in animal models, they've actually used the nanorobots to go in and destroy those. Wow. And then the nanorobots themselves disintegrate and voila, you have a clean blood vessel without the need to do arterial surgery. It's wild. But as the programming gets more and more sophisticated, you could actually begin to assemble such robots into essentially self-assembling microchips inside the person's brain without ever having to do surgery on them. Yeah, one thing that they said was pretty wild was that these robotic electrodes that they were inserting were incredibly small compared to even like a hair, a hair on our head. And they were like the size of almost a neuron.

Yeah. That's one of the things is we're getting things down to cellular or in some cases subcellular diameters and sizes, which makes them, of course, much more compatible with biological systems, because literally they can fit into the interstitial spaces between cells. And as the technology gets more and more sophisticated, and placement becomes more sophisticated, that permits greater and greater modulation of whatever the biological system is doing. In some ways, and I realize this is,

kind of a frightening thought for some people, but I think perhaps the next step in human evolution may be, to some extent, a sort of marriage of technology and biology inside ourselves. I mean, we've been a technical species for at least the 35,000 years that modern humans have existed, and certainly well before that, with Homo erectus. But our technology has thus far been mostly external. And now we're beginning to look at, well, what happens if we put the technology inside? Yeah, one thought I had about that was that, you know, like, let's say you have the first 1,000 people who put these things into their brains and then their IQ jumps up to, like,

double. So they go from like, you know, maybe an IQ of 100 to like 200, literally overnight. Elon Musk made a comment that, you know, when out of the IQ, you know, when, about the cost of it, he made a subtle comment that if you had AI integrating with you in this, you would quickly be able to pay it off. And the thought there is, like, imagine your IQ jumps to 200. And now, you know, you're a physician, for example, and you know every disease and every, you know, it's like all the textbooks of medicine are right there at your disposal in a way that they weren't

before how much more value can you bring to someone in healing their disease you know and so this is like this is going to give that person an advantage over everyone else who doesn't have that yes so that is going to drive you know this sort of um increased competitiveness that we're talking about this evolution of sorts and you can think about this just with like a smartphone like how much more productive does a smartphone make you? Well, some people make them less productive, but some people make them more productive, right? Yes.

So it's like, you know, your smartphone and you are kind of like in this symbiotic relationship for most people. Yeah, well, indeed. Most people these days, if you say, I want to take away your smartphone, you can watch the panic evolve. Oh, man.

But, you know, I mean, we're going to face a lot of hurdles and decisions, but that's been true at every stage in which we've developed new technologies, because essentially, in many ways, the technology itself is neutral. A lot will depend on how we decide to apply it, what limits we put on it, how that interacts with social structure, social justice.

In many ways, humanity right now is at a, an interesting crossroads in that we've grown as a species to the point where we now truly dominate the environment at all levels. And the choices we make will either result in a very positive future or we have the ability to perhaps screw up and do away with ourselves. Which way we will go? I'm not sure. You know, it's interesting.

A lot of people are critical of like Instagram and social media. And yet social media is created by us. You know, it is the outpouring of our id, the base drives at time, you know, sexuality, power. You know, and then there's, you know, I try in my social media to promote things that are good, you know, including, you know, empathy, how do we connect with other human beings better, that sort of thing. As a species, frankly, we're capable of great good and of great evil. I think one of the things that social media has done, it hasn't changed the nature of human beings. What it has done, though, is it has amplified that ability in both directions.

Yeah, and unfortunately, in terms of the amplification, like the people who post, the most obnoxious, most horrendous things often get a ton of comments on like Twitter, for example. So one of my recent tweets, there was this one guy who posted something about like, you know, women should stay home and be contained by the patriarchy. And, you know, he only had like maybe 10 likes on this post, but he had about 10,000 comments. And it's because, you know, anyone who saw that was just immediately enraged and was like, you know, wanting to attack this guy, what people don't realize, and this was what my post was, was that the best thing that we should do is completely ignore these types of characters and give them zero energy, because any energy that you put into them will only increase the exposure that they get.

Yeah, indeed, feeding trolls is not a necessarily good thing to do. so we're but what we're talking about is like um what we're talking about is how humans interact with technology and how technology is it's not good or bad it is it almost like exposes who we are maybe underneath and technology provides a mirror because of course the technology is created by us and then even more so the uses to which we put it are created by us and then even more so the uses to which we put it are created by us, and those can be either good or bad. A good example of that is, you know, when we were much more primitive and our only weapons were the occasional club or a rock,

well, yes, we could inflict individual damage and we could fight with each other, but we weren't capable of destroying an entire planet. We now have made weapons where that is a possibility. So on the evil side, we can go much further, and on the good side, we can go much further. A lot's going to depend on as a species whether we're mature enough to handle what we've created.

I want to dive back a little bit into the potential for, you know, we already have the deep brain stimulator. We already have some of that as an application. When you think about like something like schizophrenia, how would you see this, helping someone who has schizophrenia? In a couple of areas. One, we know some of the areas of the brain that are chronically malfunctioning

as schizophrenia. For example, the dorsolateral prefrontal cortex, the area that kind of serves as the brain's scratch pad, if you will, is hypoactive. That's thought to underlie one of the mechanisms by which people develop delusional or fixed false beliefs. because the reason that all of us don't do that is normally one of the functions of that area of the brain is to compare your current experience to past experience and essentially provide you with a probability estimate.

About 60% of the population will admit to having had at least one psychotic experience where they saw something that wasn't there, they had an illusion that they imagined something was something, else. They heard their name in a closed room. Nobody was there. For the healthy individual, their dorsolateral cortex says, okay, that doesn't quite fit what I understand about reality or any of my past experiences, so I don't think that was real. If you're, that area of your brain is not functioning, you may not be able to do that. And then you begin to develop a story or a narrative to try to explain how this happened. Well, maybe that was the voice.

of God. Maybe that was the voice of a demon. Other areas, the hallucinations themselves, appear to be a malfunction of the arqueous facet facetilus between Wernicke's area and Broca's area. A computer system might be eventually capable of detecting when that system was malfunctioning and containing it, thereby limiting the person's propensity to auditory hallucination. Yeah, I was thinking about schizophrenia and dream states, because in dreams, it's kind of similar, right? We're seeing a version of reality, and things are jumbled together, and we're not seeing or questioning in our dreams, right? Some of the sort of things that are going on. Have you ever thought about the parallels between our dream state and schizophrenia?

Yeah, and certainly in dream states, the normal rules. of logic don't apply. And that tends to be true in psychotic illness as well. And people have indeed wondered whether part of the problem with schizophrenia is a breakdown in the filters that normally keep the illogical or non-logical processing of information limited and allow it to spill over into areas where it's inappropriate. And certainly with, you know, as far as we can tell, now we don't, I think we still don't understand all of the functions of sleep all that well.

But one of the things that REM appears to be doing is it's in some ways kind of similar to defragmenting your hard drive. It goes in and consolidates and removes extraneous material. And indeed, if you REM deprive people for about five days, you wake them up every time they start to enter REM. they'll start to have breakthrough visual and auditory hallucinosis while awake. And indeed, that is similar to what you see with some people with a primary psychotic illness. Gosh, I love talking to you. So, okay, so schizophrenia, there's a couple areas that might be targeted. How about with something like obsessive-compulsive disorder?

Obsessive-compulsive disorder is attractive because, we actually understand the circuit involved relatively well. One of the functions of the dorsal striatum that includes the caudate nucleus is essentially this is where we store pre-programmed motor behaviors. Those things we don't want to have to think about. We don't want to have to dedicate cortical time or effort to it. And in all animals, including us, part of that includes grooming and orienting behaviors. When the caudate nucleus thinks we need to orient or we need to check, it will generate a signal that goes from the anterior caudate nucleus to the orbital cortex and trigger a behavioral response.

We'll go check to see that we turn the stove off or we lock the door or we'll straighten our hair. And then the behavior generates a negative feedback signal from the orbital cortex back to the globus pallidus and then up to the caudate nucleus, which is to turn that off. In OCD, the off side or the inhibitory side of that circuit doesn't work right, which then the caudate keeps going. It keeps perseverating and repeating the orient signal or the grooming signal, which is why these people get stuck in rituals. Electronically, if we were to boost the off signal,

when that negative circuit was activated, it might help fix the illness. We do that indirectly by altering the neurotransmitters, especially serotonin available in the orbital cortex, which strengthens that signal some, but usually it's an incomplete or inadequate strengthening so that the person doesn't, they're not cured of their OCD, but their symptoms become less intense. Yeah.

I think, what about, have you, have you seen some of the studies with the deep brain TMS? Yes. What are your thoughts on that? I think that's very, very positive. It opens another window of opportunity. again, it's in its infancy, and we don't yet know quite all of the things we can do with it, but by varying the number of magnets and the configuration of magnets,

we're now being able to reach deeper into the brain and affect deeper circuits. One of the limitations of the original RTMS was that you were only able to generate electrical currents down to about two centimeters below the cortex. And that's because, of course, magnetic fields drop off at the square of the distance, which is a fairly rapid drop off. But you can change the shape of magnetic fields by changing the shape of the coils in the electromagnet. And now people have gotten to the point where they can generate magnetic fields

that will reach deeper into the brain, which allows targeting of deeper circuits. Because frankly, in a lot of illnesses like OCD, it's fairly deeply located circuits that are not functioning correctly. So we may be able to, if you will, magnetically reach into some of those areas that have been beyond our reach up to this point. Okay, so we talked about schizophrenia. We talked about OCD.

What do you think about depression in particular? What areas would you specifically target or see? being targeted. Well, again, I think the limbic system and the limbic system is fairly complex. This is in the dominant hemisphere.

This involves the anterior temporal lobe, the amygdala, the hippocampus parapherapal complex. The circuit then swings up to the anterior mammillary bodies and then finally up to the anterior cingulate cortex.

There may be several points at which altering the activity of that circuit may improve mood and attention as well as energy and outlook. Certainly in cases where people have had RTMS to sort of just globally stimulate the anterior portion of that circuit, they get a pretty good antidepressant response across the number of studies, more direct influence on the circuit at more discreet points might give us even more benefit once we understand where the critical points are. Yeah, we looked at one article together a couple months ago, actually, so we've been talking about this for a while, in which through

dense EEGs on the outside of the brain, they're able to basically have a, they have a specific question and then they have several answers that a person can choose from and they're able to with a fairly high degree of accuracy be able to discern which which of those answers the brain is thinking yes um and currently in the neurolinguistic work that's been done um you know if if essentially the work they've done is they've had people imagine that they're about to see that they're about to say a word. And that sets off the motor programming, because your brain's getting ready to actually have you say the word.

If you put high-density EEG electrodes over Broca's area and the related association cortex and have the person repeat words over and over and over again in their mind, usually about 10 repetitions, the computer can learn to recognize the EEG pattern associated with that specific word. If you have the person do enough words, obviously, then they can begin to think phrases, and the computer can actually then generate via artificial voice what the person is thinking. You know, most human beings, in terms of day-to-day vocabulary, they're about a half a million words in the English language,

but we only typically use about 7 to 20,000 words. And that's easily within the data process. parameters that a computer could learn and could associate with the EEG activity. So for some people who may be due to stroke or other brain injury of loss, the ability to speak, they may eventually be able to have a computer directly speak for them. Yeah. Yeah. And they also talked about visual.

Like if someone is blind, they may be able to help them develop some sort of vision again. Yes. Indeed, one of the nice things about the primary visual cortex is its architecture is fairly well understood. And indeed, I think everyone who went to medical school remembers in histology learning about the layers of the cortex, neocortex having six and paleocortex four. Actually, that's not the way the brains organize, though. The layers are visual, if you're looking at a microscope, but actually, the neurons that make up those layers are arranged in columns, little processing units where the data goes in at the top of the column, and depending on what's happening, there may be an output that comes out the bottom of the column.

If you understand how those columns function and how they cross-communicate with each other, you may be able to then interface them with microelectrodes that will allow input from perhaps of isizing, or some other sensor device that will at least tell the person what the world visually looks like. Yeah, do you think you would go for that? I think once the technology is developed enough, yes. Yeah. One thing you may not know about Dr. Cummings is, do you want to... Sure.

Yeah, I am blind, have been for a number of years now. I suffered a choreo retinitis at 31, and that was the end of vision for me. And certainly one of the areas that may eventually become possible is restoration of vision via artificial technology. There have been some initial steps in that direction. And because the mapping of the visual cortex is fairly straightforward, I think over time, as the technology improves, the potential for visual restoration will become better as well. Another area that people have, of course, been working on is artificial limbs. Some of them can now be controlled by having the person think about contracting muscles more proximal than where the artificial limb begins. I think eventually that may get replaced with the ability to actually think about moving and have processors and perhaps an artificial leg or artificial hand actually respond.

Are there any other big psychiatric disease categories that you want to speak to in regards to, you know, potential for, you know, the more severe cases, right, to have an app for that? Yeah, I think, well, again, I think we're going to see a marriage between biological research and technology. we may eventually be able to begin to repair or replace some of the damage and losses that occur in neurocognitive disorders, which would be a major benefit, but, you know, in terms of restoring brain structure. I know the, you know, frankly, I go back far enough that the dogma at one point was an adult brain doesn't change. Well, that turned out to be dead wrong. those little satellite cells they used to talk about actually are neuroblasts and we can grow new neurons. But if we had a combination of the ability to replace neurons and the ability to perhaps artificially modulate what some other circuits are doing,

we may be able to improve things like neurocognitive disorders much more than we can now. In fact, currently we don't really have any effective treatments. with some medications slow progress, but that's about it as far as we've gotten. Yeah. Yeah, I think there's so much room to explore as well, the brain. One thing that I thought when I was, you know, reading about this and thinking about it is, you know,

up until recently when we've spoken about mere neurons, we've thought more about apes and the ape studies and or MRI studies. and the neuroscientist actually mentioned in the presentation that, you know, when you look at someone moving, your brain lights up in the same way. Yes. And so you can think about your movement. That makes your brain light up. You can make the movement.

That makes your brain light up. Or you can see someone else making the movement. And that makes your brain light up. Yes. And so the seeing someone else make the movement and your brain lighting up in a similar way, is what we call mirror neurons. And the science of that is, I think, foundational to how we understand empathy, affective empathy in particular,

and how we learn movements from watching other people make the movements. And interestingly, in this talk, they mentioned that that's one thing that they were observing in the human, or, you know, in, well, they haven't done this in humans yet. They had observed it in their studies that they had done. and so they were talking about connection and how we connect and how we speak to each other and potentially there will be another way to connect which will be brain to brain in a way that is more through technology but that our brains can speak to each other I don't know if you had any thoughts about that that's certainly a possibility I know there has been some research

albeit very preliminary in the UK where people have had implanted peripheral chips, that is chips attached to the peripheral nervous system that responded to changes in sympathetic tone. This was a husband and wife research team where if either of them got upset or angry and their sympathetic activity went up, the microchip in their arm would sense it. So they intuitively knew how the other person was feeling in a way that didn't require words or observation, which was kind of a unique experiment. And indeed, they reported that over time they felt they had become more empathetic toward each other because they had essentially a moment-by-moment reading of at least

what some emotions were going on in the other person. You know, certainly human beings do that all the time in terms of things like subtle movement, subtle facial expression. But this, I think, was the first example I had read about of microchip to microchip transfer of empathetic information. Yeah. I almost wonder like if there was an app, you know, I think someday there will be like apps for this type of thing.

And I was thinking of one app in particular, which would be really interesting, is that, you know, you basically get the ability to read people's emotions, like instantly when you turn this thing on like 100 fold better, right? So for example, most people are actually not very good at reading. micro expressions. And I've had a lot of training in this and I've given other talks on this in my podcast. And so what if instantly someone was able to see the micro expressions for what they were and then would that help them communicate better? Would that help them see the other person's

emotions in a more accurately, a more accurate way and instantly increase their social IQ? Now, I think if you're listening to this, I would highly recommend that you do this training yourself rather than necessarily, you know, needing an app for a machine brain interface to do it. But I think like, you know, when you think about like all the potential possibilities and what's going to happen, it's pretty mind-blowing. Yeah, we're living in a very exciting crossroads in human history and human evolution in terms of the interaction between ourselves. and the technologies we've created and the possibilities that may open up.

One thing they talked about as a, you know, he mentioned that potentially you could put thoughts into your brain through the, you know, the interface. So not only do, are we they're reading the thoughts, but they're putting new thoughts into your brain. And someone in the comments section said, huh, all of a sudden I got this neuralink and all of a sudden I feel very, very passionate about buying a Tesla. I'm sure there would be issues with advertising. As we've seen with things like Amazon and Google, one of the interesting social science experiments I read about was that if you have answered 100 likes or don't likes on something like Amazon or Google,

the algorithm is then better at predicting your take, than anyone in your family. Oh, wow. Well, I actually, you know, I enjoy sometimes what they'll recommend. Because it's like, yeah, I would like that. Yeah. Well, and that's the entire purpose of that particular business model with respect to targeted advertising.

And those algorithms have gotten quite good at figuring out probabilistically what it is, you will like based on the ability to look at all of your past likes and dislikes. Yeah. And for some people, that freaks them out for other people. We're just going with the ride, hoping that these companies don't turn evil on us or something. Yeah, well, indeed, it may be positive or negative. It depends on what we do with it. Right.

Right. Well, I think we need to wrap things up. Any final closing thoughts you have on this? primarily with respect to technology, it's an exciting time to be alive. I look forward to seeing what evolves over time, including what evolves with respect to our own evolution, if you will. And I think that's probably where we can leave it. Yeah, so my concluding remarks would be,

I think where I'm most excited about this progressing in the near future is for, you know, if you think about like a bell curve and there's like people who are like three or four standard deviations on the, on the mean of the bell curve, suffering because of that, you know? So for example, they've suffered with severe suicidal thoughts their whole life, or they suffered from severe OCD where even the medication, even the psychotherapy has not worked, you know, can this be another form of treatment for those very severe cases? And I hope that's where it starts. honestly it could be a really bad sci-fi horror movie or it could be really beneficial you know so i

think um having some constraints and they talked about like they would be against advertising for example but having some constraints like that um and really thinking through the ethics of the integration of these things into a common person's experience i think is a whole other exploratory realm that needs to be done. All right. Well, thanks for coming on. Thank you.